Seagate ST31000340AS 5v TVS issue

[netw3rker]

Hi Guys,

I've got a couple quick q's for you about the TVS issue on the Seagate ST31000340AS series of drives. I'm working with one now that came from an external drive that suddenly stopped working. No word on power surges or wrong power adapters etc.

My initial assumption was that it was the 0 byte lockout issue, but when i tried to get a terminal read, there was no data reported back at all. since this was my first time using that setup, i assumed it was my hardware & just ordered a bunch of new stuff (still in the mail). In the meantime i kept reading the forums here & found out about the TVS Diodes being a common problem. I picked up a multimeter from the local "cheap parts from china" store and checked the diodes. turns out the 5v TVS was definitely shorted, and the 12v TVS was fine.

I wasn't able resist the urge to follow the instructions on those forum posts and simply remove the 5v TVS, fire it up and see what happened. The drive did spin up just fine and all the sounds were normal, but the computer still couldn't read the drive. Which really leads me to my next question. shouldn't the spot where the 5v TVS was removed from be bridged somehow? I'm sorta under the impression that diodes allow current to flow past them, but only in one direction and zener diodes only let x amount of voltage flow that direction. I'm assuming that these are essentially zener diodes that are ensuring that just the right amount of power hits the board.

If that spot is supposed to be connected/bridged, what are the recommendations for gauge/wire/composition etc? I feel like "just remove the bad part" is only half the solution, so if there's any other advice please let me know!

For the experts who are going to now say "take it to a professional", your voices are heard as always. My next step is to send the board to a place that will swap the rom out to a new matching board before continuing much further.

Thanks!
-Chris

[drc]

I would suggest reading up on TVS diodes or using your multimeter to research the characteristics of some TVS on a known good drive...

[BlackST]

I would suggest also to buy a dual-tracking power supply with separate current limitation , to work with damaged drives.
It saved my life and drives countless times

[Alexii]

If the drive did not spin before and spined up after u removed TVS - good. No need to bridge nothing. U located the initial issue. If the drive spins up and makes calibration sound and u hear no clicks - great. The fact that the drive is not IDing is kinda bad. So i take it BIOS is unable to see the drive ? Are u connecting it direct to mobo or through external controller? What equipment u have in ur posetion?

[netw3rker]

drc,

that was exactly what i needed! I think the first few paragraphs in the explanation here: http://en.wikipedia.org/wiki/Transient_ ... sion_diode say it all. so essentially those diodes only active when the current is over the 'dangerous' level & reroute it away from the circuitry. that would make bridging the path a really bad idea unless you like electrical smoke right?

I guess even a little homework goes a long way.
-Chris

[BlackST]

They do not exactly act like how you said. In fact guess why sometimes they work sometimes no....

[drc]

drc,

that was exactly what i needed! I think the first few paragraphs in the explanation here: http://en.wikipedia.org/wiki/Transient_ ... sion_diode say it all. so essentially those diodes only active when the current is over the 'dangerous' level & reroute it away from the circuitry. that would make bridging the path a really bad idea unless you like electrical smoke right?

I guess even a little homework goes a long way.
-Chris

Well, one end of it is on the voltage supply and the other end of it is on ground, so 0 resistance = infinite current = infinite heat etc etc.

[netw3rker]

Alexii, ATM i'm connecting via a usb->sata adapter. I'm not at all confident that it actually works so my next step is to connect directly to a motherboard that i know both works and supports drives of that size.

BlackST, I'm game for that purchase Do you have recommendations on particular products?


All, I'm a relative newbie to doing serial/ttl work and once i get the board to a working condition, i'd like to check the terminal read out just to be sure everything is working. I've pretty much been working off of this tutorial: http://www.msfn.org/board/solution-seag ... 09f8962792 which i found a link to on this site.

That seems like a lot of hardware being connected together too. does anyone have any recommendations for a more all-in-one setup/item? or is that the best case scenario?

[Alexii]

ur next step is connecting to mobo and taking a look if BIOS can see it. Also listen for clicking noises from the hdd , listen if the heads move , does it calibrate etc.

[netw3rker]

Thanks Alexii. I've dealt with a few bad HD's already and made a ton more. I can say that there is definitely no clicking noise, the heads are moving fine and it does seem to calibrate successfully. My guess is that its currently just in lockout mode now. I'll hook it up to a pc & see what bios says & update you guys.

-Chris

[netw3rker]

As a final update here, I did get the drive up and running again! @alexii, initially the drive did not post to bios (as i had expected). Once I finally got the TTL connection going, I found that it was in the BSY state. going through the steps listed in the link I sent earlier unlocked the drive & got everything running again.

I did have one minor heart attack when verifying the drive. i booted up into windows and there was no drive letter assigned. Thinking this was now the 0byte problem (even though bios now showed a 1TB disk) I pulled up the hardware list and the disk manager. Windows was reporting that there were no partitions allocated and 100% of the 1TB of data was unallocated. *luckilly*, before doing anything drastic, I double checked with my friend and confirmed that it was a backup drive formatted for his mac. whew!

FWIW, the instructions in the link I posted are missing a couple of details, namely that when you use a cr2032 battery to power the ttl bus (instead of using the PC power supply's 5v power), you have to also connect the ground pins together since TTL needs a common ground.

Thanks for all of your help guys!
-Chris

[pcimage]

Well done!!!

[dobrevjetser]

Luckily someone with brains. Good job.